Gravity
Gravity
Gravity
Gravity is remarkable. An apple falls out of a tree- what made it accelerate?
Nothing touches it. Saying the motion is due to “gravity” doesn’t explain it.
Einstein, in his general theory of relativity (1917) went very far in explaining why
there is gravity. But let’s follow Newton’s path, and merely try to describe it...
Gravity acts at a distance. Since the acceleration due to gravity is constant, gravity
apparently acts at the top of trees as well as the bottom. How high does it reach?
In Australia, gravity still acts towards the center of the earth. It appears that it is the
Suppose it travels so far, that you finally notice the earth is not flat:
This was Newton’s epiphany: the moon’s motion is of the exact same nature - due
Why not up to the moon? Why not further? Newton realized that there is one
universal law of gravity, affecting all masses. One law that could explain and unify
By 1650, the distance to the moon was well known: about 60 R e (60 earth radii).
Conclusions: the moon is about 60 times further from the Earth’s center than us,
and a(moon) = (1/60^2)*g (whereas a(for any body near the earth) = g)
F(grav) on any object here, mg, is proportional to the mass. Newton III says the
force should be equal (and opposite) on both objects that gravitationally attract: the
formula for gravity must be symmetric in the masses. Newton put this all together:
G M1M 2
Newton’s Universal law of Gravity: F12 = F21 = Fgrav =
r2
G was not measured for > 100 years, by the way, until Cavendish (1798)
Phys 1110, 9-3
the exact same as if all the sphere’s mass were concentrated at its center. (Newton
That’s a small force. The weight of a mosquito is about 1000 times more than that.
Example: What is the force that the earth applies to an object with mass m?
Oh of course! The answer is mg, just as we’ve been using all semester.
difference. Even at the top of Mt. Everest, taking r= 6.4E6 m + 8E3 m, the answer
is still unchanged to two sig figs. At the orbital height of the Shuttle,
h=250 km above the surface, the appropriate r for the formula (to find F(grav) on
Kepler was 70 years before Newton. He discovered some patterns in the data of
Tycho Brahe (this was before telescopes.) He summarized the data with the
(and it’s the same constant for all the planets around the sun.)
Newton was able to show that these laws arise directly, mathematically, from his
universal law of gravitation. (Compelling proof that his law of gravity was correct.)
Example: Let’s prove Kepler’s third law for the special case of circular orbits.
where m = mass of the planet, and R = radius of the planet’s orbit around the sun.
What supplies this F_net? What’s the physical force on the planet?
G Msun m
There’s only gravity acting, so F_net = Fgrav = .
R2
Phys 1110, 9-5
Combine these, and notice that m cancels out. Thus, no matter what the mass of the
v 2 G Msun G Msun
planet is, we will have = 2 , or if you prefer, v = R
.
R R
If you have an orbit around something else, replace M(sun) with M(central object).
(4π 2 R2 / T 2 ) G Msun
Remember that v = 2 Pi R/T, plug this in the 1st Eqn: = .
R R2
R 3 G Msun
Now rearrange: = . This is Kepler’s third law. The right side is a
T2 4π 2
constant for all planets. Newton has proven, and generalized, Kepler’s third law.
Earth satellites (the moon, communications satellites, whatever) must also satisfy
R^3/T^2=constant, but with a new constant (use M_earth on the right side now.)
Conclusion: any satellite in orbit at some radius has a certain, definite, predictable
period (and thus, velocity), independent of mass. We’re doing rocket science here!
Example: Consider low earth orbit (e.g. the shuttle) so R=R(earth) + a little bit.
G M earth
v= ≈ 8km / sec ≈ 5mi / sec. Just under 20,000 mi/hr.
R
Any slower and the shuttle will not be a circular orbit, it will fall. Any faster, and it
will be go to a higher orbit. The period (time for the shuttle to go around the earth
If you could throw a baseball (and have neither friction nor mountains get in the
for the baseball and the shuttle, because R is measured from the earth’s center.)
Phys 1110, 9-6
T is often easy to measure. (The period of the earth is one year, the period of the
moon is 27 days, etc). So you can use Kepler’s third law to deduce R. For the earth,
we know R(from the sun) and T: we can figure out the Sun’s mass. Cool!
When Cavendish measured G in 1798, he used the distance and period of the moon
to deduce the mass of the earth. Cavendish weighed the earth! (Knowing the weight
and size of the earth, we could then deduce our average density, and learn e.g. that
Example: Consider a satellite whose orbital period is exactly 24 hours. Think about
this - the earth rotates once in 24 hours, and in that time the satellite has also run
around the earth once. From our perspective on the ground, that satellite is at rest
always above the same spot. You aim your dish at it and don’t have to keep tracking
the satellite. We call this geosynchronous orbit.. How high up is such a satellite?
G M earth 2
R3 = T . Plugging in all the constants, with T=24 hrs, and taking the
4π 2
cube root (figure out how to do that on your calculator) I get R=42,000 km.
This is a very high orbit. Since R_earth = 6,000 km, it is 36,000 km above the
shuttle. The farther out you are in orbit, the slower you go. (Bigger distance, but
still bigger T.) Puzzler: Do you think the moon’s speed is larger or smaller than 3 km/sec?
Phys 1110, 9-7
Energy considerations.
F=G M1 M2/r^2, not mg. The dot product has a “cos θ ” in it, as usual. It’s subtle:
180 degrees, cos θ = -1, so this cancels the - sign in the formula.
rf G M 1M 2 1 rf
ΔU = + ∫ dr = G M1M 2 (− )
r0 r2 r r0
$ 1 1'
U(rf ) − U(r0 ) = G M1M2 & − )
% r0 rf (
From this, we can argue that Ugrav (r)= −G M1M2 / r . (NOTE: 1/r, not 1/r^2)
Think about that - sign, it’s there to give back the formula we just derived for Δ U.
This definition of U(r) says the potential energy is zero when r gets very large.
A reasonable choice: far away objects feel nothing, no force, no potential energy....
(You can make other choices, the zero of U(r) is arbitrary, but this one is common.)
According to this formula, any object (closer than infinity) has a negative potential
energy, meaning you would have to ADD energy to it to move it away to infinity,
where it’s defined to have zero potential energy. (Like a ball in the basement,
which we said has negative potential energy, if we define U=0 at ground level.)
Phys 1110, 9-8
U(r)=-GM1M2/r.
U(r)
r
This graph is called a “potential well”.
(As you move away from the earth, your potential energy is higher.)
I drew it symmetrical to remind you that the variable r is radius, from the center.
You can head away from the earth in any direction you want, U(r) rises as you go
off any which way. U(r) depends only on radius: the potential energy 20 feet off the
ground is the same no matter where around the sphere you are.
Escape velocity: If you throw a rock straight up it goes up, stops, then comes down.
Think of energy conservation - total energy is conserved: it starts off all kinetic,
turns into gravitational potential, then turns back into kinetic on the way down.
What if you throw it very hard? Can it ever escape, and never return?
The answer is yes. You need to give it a critical amount of initial speed, called the
If v0 =v_esc, then the rock will JUST barely make it off to r= ∞ : it will have zero
energy out there (K=0, U=0, E_tot=0) But energy is always conserved: in this
critical case, the initial total energy must have also been zero. Let’s write this out:
Phys 1110, 9-9
1 GMearth m
For any object near the surface, E tot,0 = K0 + U0 = mv20 − .
2 R earth
If we equate this to E(final)=0 (to find the critical initial, or escape, velocity)
2GM earth
v esc = .
R earth
m has cancelled out. This is Sqrt[2] times bigger than the low-earth orbital velocity
we found on p. 5. Only 40% more initial speed will let the shuttle escape earth!
If you throw the rock even harder, it has greater initial energy. It will escape, and
still have some positive energy off at infinity, i.e. it will have K>0 far away.
Graphically, you can think of this like our old roller coaster pictures:
with.
For the situation shown above, the sum is >0, you’re “free”, you will ultimately
If the sum is negative, you’re “bound”, you will never reach infinity. (That’s the
case for all of us at the moment, and even the astronauts in the space shuttle.)
Phys 1110, 9-10
An object in circular orbit does not have Etot=0. It’s in orbit, it’s not running off to
1 GMm
For objects in circular orbit, E_tot =K+U= mv2 − .
2 R
Using our earlier formula for v(orbital) (see notes, p 5), I find K= GMm/2R.
This is definitely all a little hard to understand at first, but it does make some sense.
• K and U are intimately related for circular orbits (because v and R are related).
• The faster it goes (bigger K) the more negative E_tot is, that means it has less
total energy. Yikes! Can that be? Faster means less energy? Yes, we just saw that
objects in orbit go faster if they are closer to the earth. Closer objects are faster.
They have LESS overall energy (they’re lower down), but have more KE .
Remember energy is the sum of K+U. In this case, as you go further out, K goes
down, U goes up, but U wins (because of a curious factor of 2 in the equations)
All this weirdness is very real. When the space shuttle pilot wants to catch up to a
satellite, she does NOT hit the gas (i.e. fire rockets directly aft). That would
increase their energy => move them to a higher orbit. More energy, but less K, so
they would watch the satellite go below them, and also start to move ahead of them!
Very strange indeed. (Shuttle pilots need to study a lot of intro physics.)
Phys 1110, 9-11
Our new formula for gravitational potential energy, U(r)=-GM1M2/r, seems way
different from the old formula U(h) = mgh. How can we reconcile them?
U(r)
If you zoom in on that graph of U(r)=-GM1M2/r, near Re r
on U(r) near the surface. But, how does it work out quantitatively? Just let
GM e m
r = R e + h , and plug into our formula U(r) = − , to get
R
GM e m GM e
U=− =− h m
.
Re + h Re (1 + )
Re
1
It’s a very useful mathematical fact that ≈ 1 − ε , if ε is small.
1+ε
Here, h/Re is small (that’s the assumption that we’re near the surface: the height is
GM e h GMe
U≈− (1 − )m = constant + hm
Re Re R 2e
The value of the constant is irrelevant. Remember, you can always choose wherever
you want to call U=0. (Adding a constant to U(r) doesn’t change physics. All you
care about is Δ U, and our constant cancels out when you subtract) So the formulas
look different, but they are in complete agreement if you are near the earth's surface.